Floneum is an open-source platform for building AI-powered workflows using large language models through a visual and extensible interface. The system allows users to design complex AI pipelines using a drag-and-drop workflow builder rather than writing extensive code. It focuses on enabling developers and researchers to create language model applications that combine different tools, data sources, and AI capabilities into automated workflows. Floneum supports a plugin architecture that allows external components to extend the platform while maintaining isolation and security. Many plugins can be written in different programming languages and compiled to WebAssembly modules, allowing them to run safely within the system. The platform is implemented primarily in Rust and emphasizes performance, modularity, and local execution.
Features
- Visual drag-and-drop builder for AI workflows
- Plugin system supporting WebAssembly modules
- Local execution of AI pipelines for privacy and control
- Language-agnostic plugin development support
- Modular architecture for combining multiple AI tasks
- Extensible pipeline framework for LLM-based applications